Tanulmány
György Rákosi
Sightseeing around our own possessive adjectives in Hungarian
*Abstract
The paper describes the grammar of the Hungarian possessive adjective saját ‘own’ in comparison to its English counterpart own. Both items can function as possessive intensifiers, but saját ‘own’ in Hungarian also has a productive non-possessive use. In addition to the basic possessive adjective saját ‘own’, Hungarian has two further, slightly archaic but still productive possessive intensifiers: tulajdon ‘own’ and önnön ‘own’. The paper draws up an inventory of the determining grammatical features of these items, and it argues that they instantiate slightly different strategies of marking emphatic possessive relations.
Keywords: anaphora, intensifier, possessive structure, possessive adjective, reflexive
1 Introduction
The Hungarian emphatic possessive adjective saját ‘own’ is a frequent translational equivalent of the English possessive intensifier own. Just like its English counterpart, it can be added optionally to possessive constructions (1b), thereby emphasizing the nature of the possessive relation in a manner similar to the meaning contribution of own.
(1) a. Ez az én vélemény-em.
this the I.NOM opinion-POSS.1SG
‘This is my opinion.’
b. Ez a saját vélemény-em.
this the own opinion-POSS.1SG
‘This is my own opinion.’
The descriptive literature on Hungarian focuses on the categorial identity of this element, suggesting that it is primarily an adjective but it can also behave in certain ways as a personal pronoun (see Keszler 2000: 162). In (1b), for example, saját ‘own’ may prima facie appear to occupy the same position as the overt pronominal possessor in (1a). It is noted occasionally that two other, slightly archaic but productive forms, tulajdon ‘own’and önnön ‘own’, also have a similar function.
* This research was supported by the European Union and the State of Hungary, co-financed by the European Social Fund in the framework of TÁMOP-4.2.4.A/ 2-11/1-2012-0001 ‘National Excellence Program’.
In this paper, I present an overview of the grammar of these three elements. I focus on saját ‘own’, the most frequently used emphatic possessive adjective. I argue that the categorial issue, in and of itself, foregrounds only some of the interesting properties of possessive adjectives, and the three elements discussed here each instantiate slightly different grammatical strategies of modifying the underlying possessive relation. I refer to saját ‘own’
as a possessive adjective in line with the common practice in the literature (see, for example, König & Vezzosi 2008 and Alexiadou 2005 for overviews). This terminological practice, as we will see, receives more empirical support on the basis of the Hungarian data than in the English case, but the more interesting questions lie beyond the issue of the categorial identity of these elements. I will argue that saját ‘own’ has a wider array of uses than own, whereas tulajdon ‘own’ and önnön ‘own’ are more restricted not only in the stylistic sense, but also with respect to the grammatical environments in which they can occur. The empirical data that I discuss in this paper have been mostly collected from the British National Corpus (Davis 2004-) and the Hungarian National Corpus (Váradi 2002).
The structure of the paper is as follows. In Section 2, I provide a thorough overview of the grammar of saját ‘own’, and I argue that saját has both a productive possessive use and a non-possessive use in Hungarian, which makes it different from the English intensifier own.
In Section 3, the emphatic possessive adjectives tulajdon ‘own’ and önnön ‘own’ are studied in the light of the previous discussion on saját ‘own’. I Section 4, I conclude the paper with a brief summary and an outlook on remaining issues.
2 The grammar of saját ‘own’
2.1 On the known history of the word
The word saját ‘own’ occurs first in written Hungarian texts in the early sixteenth century. It appears to have had roughly the same function as it does in contemporary Hungarian even in this early age of its known history, but its etymology remains unknown (see D. Mátai 1999, Bárczi 1941, Zaicz 2006). It does have, however, several derivatives with first occurrences mainly from the late eighteenth and early nineteenth centuries. I list some of these in (2), indicating the first occurrences in brackets as given in Zaicz (2006: 714) and (Bárczi (1941:
264):
(2) a. saját-ság
own-NOMINAL.SUFFIX
old meaning: ‘possession’ (1577)
modern meaning: ‘characteristic feature’ (1836) b. saját-os
own-ADJECTIVAL.SUFFIX
old meaning ‘possessor’ (1780) modern meaning: ‘peculiar’ (1832) c. saját-ság-os
own-NOMINAL.SUFFIX-ADJECTIVAL.SUFFIX
‘peculiar’ (1838) d. saját-ít
own-VERBAL.SUFFIX
‘disposses’ or ‘acquire’ (1834)
Thus items in this family all denote some aspect of possession, or they have the shifted and historically apparently later meaning ‘peculiar/characteristic feature’.
Own is a derivative of the Old English agen, which is the past participle of agan ‘own, possess’ (König & Vezzosi 2008). König & Vezzosi (2008) carefully show how the Old English participle has undergone a process of grammaticalization, and they also argue that the cognate German eigen ‘own’ or the Italian proprio ‘own’ followed a similar path of development into a status more functional in nature than what their historical ancestors used to have. Unfortunately, such a historical path of grammaticalization cannot be documented for Hungarian given the lack of data from periods in which saját ‘own’ behaved in a manner significantly different from its contemporary use. I argue below that saját ‘own’ is not as strongly functional in nature as own in English since it can occur in a wider array of grammatical contexts. This could be an indication that saját ‘own’ in Hungarian is less grammaticalized than own in English, but the lack of sufficient diachronic data for Hungarian does not automatically warrant such a conclusion (even if cross-linguistic considerations may make it plausible). For this reason, I concentrate below on the synchronic grammar of saját
‘own’.
2.2 Two syntactically distinct uses of saját ‘own’
The English own has two characteristic syntactic features: it co-occurs with a possessive determiner and it precedes any adjectives that modify the possessum (see König & Vezzosi 2008). The following examples from the British National Corpus (Davies 2004-) illustrate these features:
(3) a. Each stood separately in its own little suburban garden.
b. Of course Boggers’ isn’t the only teacher with his own funny little ways; …
This strong relation between the possessor and own has lead some to assume that one’s own is a complex pronominal element in English (see especially Zribi-Hertz 1995).
Nevertheless, König & Vezzosi (2008: fn. 1) note that the expression an own goal does exist in contemporary English, but they consider it a relic of an earlier usage, which is not productive any more. That this use is indeed syntactically distinct from the regular possessive use is shown by the fact that evaluative adjectives precede own in this case (4a). A few other non-possessive own+noun expressions that have a lexicalised character can still be found in contemporary English texts (5). All the examples in (4) and (5) are from the British National Corpus.
(4) a. The game had everything - two spectacular own goals, brilliant strikes and marvellous individual skills.
b. There could also be an own band product in the very near future.
(5) a. Well I've got my nice own bicycle.
b. They've got no thought only for their selfish own ways you see.
The two examples in (5) are different from the ones (4) in the sense that own bicycle and own ways do not appear to be lexical units with a distinguished meaning, unlike own goal, own brand or own resources. However, it is remarkable that both sentences in (5) are from the same source of recorded conversations from the fourth quarter of the twentieth century. The
construction itself is very rare, and it only has a few occurrences in the corpus. Thus we can safely conclude with König & Vezzosi (2008) that it is indeed not present productively in contemporary English, even if it used to be available in earlier periods of the English language, a fact which König and Vezzosi amply illustrate.
In Hungarian, saját ‘own’ also dominantly functions as a possessive intensifier which precedes any other adjectives, including evaluative adjectives. (6a) is from the Hungarian National Corpus (see Váradi 2002), and (6b) is also a corpus example simplified for the purposes of this paper:
(6) a. Meixner Mária a saját kis élet-é-t építgette.
Meixner Maria the own little life-POSS.3SG-ACC built
‘Meixner Maria was working on her own little life.’
b. Nem beszélt a saját esetleges újrajelölés-é-ről.
not talked the own potential renomination-POSS.3SG-about
‘He did not talk about his own potential renomination.’
Besides this core possessive use, saját ‘own’ can also occur in non-possessive constructions.
Occasionally, it forms a lexical unit with the following noun, as happens in the case of sajátfüggvény ‘eigenfunction’ or sajátnyelv ‘idiolect, own language’, where the standard single-word orthography itself is already an indication that a lexical word formation process has taken place. However, there are numerous examples in the Hungarian National Corpus in which non-possessive saját ‘own’ combines with a noun in what appears to be a productive manner. Consider (7):
(7) a. Szó van egy új saját lemez-ről … word is a new own album-about
‘They are talking about a new album of their own …’
b. Jó saját kezdeményezések is vannak a települések-en, … new own initiatives too are the communes-on
‘The communes also have their own good initiatives, …’
c. … inkább a rosszabb minőségű saját gaboná-t vetik el a föld-be.
rather the worse quality own corn-ACC plant PRT the ground-into
‘… they rather plant their own worse quality corn into the ground.’
What these examples share is that saját ‘own’ is preceded by an evaluative adjective (or an attributive participial phrase in (7c)), and the noun head it modifies is not possessive. These examples can only be rendered in English using some possessive paraphrase, but note that the Hungarian examples are not possessive marked.
This non-possessive construction is evidently productive in Hungarian, and it is not restricted to specific lexical combinations. The Hungarian National Corpus includes a great number of instances of the non-possessive saját ‘own’ construction, the list of nominal collocates including such nouns as bank ‘bank’, büdzsé ‘budget’, cég ‘company’, elképzelés
‘idea’, elnök ‘president’, földterület ‘land, area’, gázforrás ‘gas resource’, idő ‘time’, kollégium ‘dormitory’, ruházat ‘clothing’, vagyon ‘property’, etc. What is common in all these examples is that they occur in contexts where a discourse contrast is implied between individually possessable instances of a natural kind and instances of the same natural kind not intrinsically owned by an implied possessor. Consider (8) for illustration:
(8) A mesterséges haj nem jobb, mint a saját.
the artificial hair not better than the own
‘Artificial hair is not better than one’s own.’
This context creates a contrast between natural and artificial hair, the former being described as one’s own. This construction testifies the syntactic nature of these saját+noun combinations, inasmuch as the possibility of contrasting saját ‘own’ with a contextually appropriate adjective implicating the lack of non-intrinsic ownership is a generally available option for the examples in (7), too.
Thus saját ‘own’ has two syntactically different uses, both of which are synchronically productive. Possessive saját occupies a position high in the possessive noun phrase, preceding any possible adjectival modifiers of the noun head. Non-possessive saját functions as a classificatory adjective, occupying a lower position closer to the head noun and following evaluative adjectives. Non-possessive saját is evidently an adjectival element. In the following subsections I argue that possessive saját ‘own’ is also best regarded as some sort of an adjective, albeit of a more functional kind.
2.3 Further remarks on the morphosyntax of possessive saját ‘own’
It is a well-known property of the Hungarian noun phrase that pronominal possessors agree with the possessum, and they can be pro-dropped under agreement with the noun head (see Szabolcsi 1994 and É. Kiss 2002 for general overviews, as well as Laczkó 1995 for a non–
pro-drop analysis). Pronominal possessors are spelled out overtly if they receive a discourse function (see Rákosi 2014 for a discussion of the basic facts):
(9) a. Ez a vélemény-em.
this the opinion-POSS.1SG
‘This is my opinion.’
b. Ez az én vélemény-em.
this the I.NOM opinion-POSS.1SG
‘This is my opinion.’
In the light of (9), it looks evident that saját ‘own’ does not function as a pronominal of some sort even in the absence of an overt personal pronoun possessor. Rather, it is either a modifier of an overt (10b) or a covert (10a) pronominal possessor. Compare:
(10) a. Ez a saját vélemény-em.
this the own opinion-POSS.1SG
‘This is my own opinion.’
b. Ez az én saját vélemény-em.
this the I.NOM own opinion-POSS.1SG
‘This is my own opinion.’
I discuss the nature of this modification in the next subsection. Here I present further arguments against treating saját ‘own’ either as a personal pronoun or as a reflexive.
First, while personal pronominal possessors must co-ocur with a definite article, the presence of the definite article is optional by saját if the possessor is pro-dropped:
(11) a. Ez *(az) én vélemény-em.
this the I.NOM opinion-POSS.1SG
‘This is my opinion.’
b. Meixner Mária (a) saját kis élet-é-t építgette.
Meixner Maria the own little life-POSS.3SG-ACC built
‘Meixner Maria was building her own little life.’
For a second argument, consider the following three sentences, the first two of which are quoted from Rákosi (2014):
(12) a. A magam arc-á-t kerest-em a gesztus-á-ban. Te is.
the myself face-POSS.3SG-ACC looked.for-1SG the gesture-POSS.3SG-in you too 'I was looking for my own face in his gesture. You too.'
→ You searched for your own face in his gesture. binding
b. Az én arc-om-at kerest-em a gesztus-á-ban. Te is.
the I face-POSS.1SG-ACC looked.for-1SG the gesture-POSS.3SG-in you too 'I was looking for my face in his gesture. You too.'
→ You searched for my face in his gesture. coreference
c. A saját arc-om-at kerest-em a gesztus-á-ban. Te is.
the own face-POSS.1SG-ACC looked.for-1SG the gesture-POSS.3SG-in you too 'I was looking for my own face in his gesture. You too.'
→ You searched for your own face in his gesture. binding
In Hungarian, unlike in English, reflexives can function as possessors. They license a binding reading in this capacity, which can be evoked in elliptical contexts (12a). Notice that magam
‘myself’ triggers 3SG agreement on the possessum, and so does each item of the reflexive paradigm. I argue in Rákosi (2011) that this is a consequence of the fact that the Hungarian reflexive is a grammaticalized possessive construction. The overt personal pronoun possessor in (12b) triggers regular agreement with the head noun, and it only licenses a coreference reading, which is clear from the only grammatical interpretation of the elliptical construction.
If we insert the modifier saját ‘own’ in (12b), as happens in (10b) above, the same facts remain: only the coference reading is allowed. In (12c), which includes a pro-dropped possessor and saját ‘own’, we have 1SG morphology on the possessum, and the bound variable reading is grammatical. Thus the possessive adjective saját ‘own’ is different from the reflexive possessor (no constant 3SG agreement) and it is also different from the overt personal pronoun possessor (bound variable interpretation licensed). This warrants the conclusion that I aim to substantiate here, namely that possessive saját ‘own’ cannot be treated either as a personal pronoun or as a reflexive.
That it is not a reflexive is also strongly motivated by the fact that saját ‘own’ can in fact modify the reflexive. (13) is an example from the Hungarian National Corpus.
(13) … én vagyok saját magam tanító-ja.
I am own myself teacher-POSS.3SG
‘I am my own teacher.’
Whereas saját ‘own’ follows personal pronoun possessors (10b), it premodifies reflexives.
This is yet another synchronically available reflex of the possessive nature of the reflexive itself, as I argue in Rákosi (2011). This is expected if possessive saját ‘own’ is an adjective, not a pronominal of some sort: qua an adjective, saját can modify the grammaticalized possessive structure of the reflexive magam ‘myself’. The result is an emphatic reflexive possessor.
A final argument in favour of the adjectival nature of saját ‘own’ comes from the fact that possessive saját can occasionally accept degree modification. In the following examples from the Hungarian National Corpus, saját ‘own’ is in the superlative (14a) or is modified by the degree adverbial nagyon ‘very’ (14b).
(14) a. Test-ünk a legsajátabb magántulajdon-unk.
body-POSS.1PL the most.own private.property-POSS.1PL
‘Our body is the most personal private property of our own.’
b. …szeretik… nagyon saját, de fogyaszthatalan produktum-a-i-t.
like.3PL very own but inconsumable product-POSS.3SG-PL-ACC
‘They like his very own but inconsumable products.’
The two examples may arguably have a mildly coercive character, but this is due to the fact that saját ‘own’ does not denote a naturally scalar property. The superlative form nevertheless occurs 83 times in the Hungarian National Corpus, and degree adverbial modification is also well attested on the web. I regard this possibility as a further manifestation of the adjectival nature of saját ‘own’.
2.4 Possessive saját ‘own’ as an intensifier
The evidence presented in the previous subsection supports the adjectival analysis of the possessive saját ‘own’ in Hungarian. Thus the emerging picture is that this item is always an adjective, but it can be inserted in two different positions in the Hungarian noun phrase. It can occupy a low position among the modifiers of the noun-head, following evaluative adjectives.
This use does not require the presence of a morphosyntactically marked possessive relation.
That non-possessive, low-level instances of saját ‘own’ are instances of an adjectival use does not require much explanation.
The morphosyntactic evidence for the adjectival status of possessive saját ‘own’ seems undisputable on the basis of the data surveyed in 2.3. The fact that saját ‘own’ is less constrained in its syntax than its English counterpart own also points towards a general adjectival analysis. Nevertheless, it has been repeatedly claimed that his own and its peers are referentially more dependent than his, and that the English own is a reflexive pronominal of some sort (see Fiengo & Higginbotham 1981, Quirk et al. 1985, Chomsky 1986, and subsequent literature). Examples of the following kind indeed seem to suggest that his/her own requires a local antecedent (Nishiguchi 2009):
(15) a. Maryi used heri/Sue'sj pen.
b. Maryi used heri/*Sue'sj own pen.
The intuition behind the reflexive analysis of own - whether it is assumed to form a complex with the possessor or not - draws on data of this sort. However, an alternative line of research
has uncovered an abundance of data that show that there is no strict syntactic constraint on the antecedent of one’s own-possessors (see Reinhart 1983, Baker 1995, and, especially, Zribi- Hertz 1995). Since much the same conclusions apply to the Hungarian data, I illustrate with three examples from the Hungarian National Corpus.
(16) a. A saját gondolat-a-i-m válaszolnak a saját gondolat-a-i-m-ra.
the own thought-POSS-PL-1SG reply the own thought-POSS-PL-1SG-onto
‘My own thoughts reply to my own thoughts.’
b. A saját kutyá-nk oda se figyel, ha utasít-om, … the own dog-POSS.1PL PRT not listen if command-1SG
‘Our own dog does not even listen if I give orders to it, …’
c. Megsértett-em a barátodat.
insulted-1SG the friend-POSS.2SG-ACC
A saját ház-á-ban útszélien viselkedt-em.
the own house-POSS.3SG-in vulgarly behaved-1SG
‘I insulted your friend. I behaved in a vulgar manner in his own house.’
What is common to these examples is the lack of a c-commanding antecedent within the clause. In (16c), the antecedent is mentioned, but it is external to the sentence containing saját
‘own’. In general, the antecedent has to be available at the discourse level for this construction to be licensed: it is a participant whose perspective determines the piece of discourse within which the possessive adjective saját ‘own’ is expressed. In (16a), this discourse antecedent is the speaker, in (16b), it is the speaker plus an associate, and in (16c) it is the friend whose mental state the first sentence describes. Therefore I conclude with the above authors (especially Baker 1995 and Zribi-Hertz 1995) that possessive adjective constructions are anaphoric only in the extended, discourse-linked sense of the word.
This is a major motivation why some authors reject the pronominal analysis of own altogether. König & Vezzosi (2008: 189), for example, argue for an analogy between the following two constructions:
(17) a. We were even more surprised at the old man’s own death.
b. We were even more surprised by the death of the old man himself.
In (17b), the reflexive himself is an intensifier adjunct that modifies the noun phrase that immediately precedes it. Likewise, own is also an intensifier in (17a). König & Vezzosi (2008) regard the two constructions as functional equivalents, and they treat own as a specialised adnominal intensifier in English.
Safir (1996) and Nisiguchi (2006) spell out the intuition behind the intensifier analysis by proposing that own introduces a two-place relation. In essence, they claim that the adjective own inherits the argument structure of the verb own. This is a diachronically well-motivated assumption in English (see 2.2). The two arguments that the intensifier own takes are the possessor (expressed as the genitive determiner) and the possessum (the head of the possessive noun phrase). The intensifier itself establishes a predicative relation between the possessor and the possessum, to the extent that it asserts a unique possessive relation between the two. In other words, it directly spells out the existence of a designated possessive relation, which is not otherwise entailed by the possessive structure itself. Since the basic semantics of the Hungarian possessive intensifier saját ‘own’ is no different from this, I illustrate this
analysis with the following Hungarian examples (based on Nisiguchi's relevant English examples):
(18) a. Kati az autó-já-val jött.
Kate the car-POSS.3SG-with came
‘Kate came with her car.’
b. Kati a saját autó-já-val jött.
Kate the own car-POSS.3SG-with came
‘Kate came with her own car.’
c. Kati saját autó-val jött.
Kate own car-with came
‘Kate came with a car of her own.’
(18a) is a possessive structure with a pro-dropped possessor, (18b) is the same structure with the possessive intensifier adjective saját ‘own’ added, and (18c) contains the non-possessive saját ‘own’. The three sentences are not equivalent semantically. (18a) can be true if Kate borrowed or stole the car for the occasion, that is, when no permanent possessive relation is denoted or when there exists no true possessive relation at all in the real world. This interpretation is compatible with the vague semantics of the possessive construction. (18b) or (18c) would not be true in the self-same situation, for the presence of saját ‘own’ requires a more intrinsic relation between possessor and possessum. This relation is a true possessive relation in the default case. Notice that the adjective contributes this relation irrespective of the presence or the absence of the grammatical possessive construction, and there is no difference between (18b) and (18c) in this respect.
There is, obviously, much more to be said about how the insertion of saját ‘own’ changes the meaning of the possessive construction, and how it affects the referential properties of the overt or covert pronominal possessor that it combines with. These issues, however, lie outside of the scope of the current inquiry. What I have wanted to establish here is that (i) the possessive adjective saját ‘own’ is indeed an adjective, which (ii) can occur both in possessive and non-possessive constructions, occupying respectively a higher or a lower position among the modifiers of the noun phrase, and that (iii) it acts as a two-place predicate introducing a strict possessive relation in both positions. The possessive use of saját ‘own’
can be conveniently referred to as an intensifier, following König & Vezzosi (2008). This term emphasizes the functional nature of this element, which, as I have argued here, is an adjective in categorial terms. In the next section, I briefly discuss two other possessive adjectives which may also be used in Hungarian for the same function.
3 Two further possessive adjectives 3.1 A brief diachronic overview
Hungarian has two further possessive adjectives besides saját ‘own’. These are tulajdon and önnön, both of which can be translated into English as ‘own’. Both are marked in the stylistic sense, and especially the latter has a pronounced archaic character. Nevertheless, both of them are still used in possessive constructions as intensifier adjectives, and can replace saját ‘own’
in this capacity, as (19) illustrates:
(19) Kati a saját / a tulajdon / az önnön kez-é-vel írt level-et.
Kate the own the own the own hand-POSS.3SG-with wrote letter-ACC
‘Kate wrote a letter with her own hands.’
The three intensifiers are semantically equivalent in (19), and the sentence is true under the same conditions with the choice of either intensifier.
Tulajdon is a complex word with a morphologically opaque internal structure. It is derivative of the word tulaj ‘alone, in himself’, which is first documented in 1519 but which has been lost since then from the Hungarian language (see Zaicz 2006: 870, Bárczi 1941:
319). Tulaj itself probably derives from a Finno-Ugric root, but nothing certain is known about its diachronic development. Tulajdon was in common use by the beginning of the fifteenth century, and it functioned as a possessive intensifier alongside with saját (D. Mátai 1999).
The diachronic development of önnön ‘own’ is better known than either of the other two possessive intensifiers (D. Mátai 1999). Its stem is the 3SG personal pronoun ő ‘he’, to which the emphatic pronominal suffix -n was added. The reduplication of this suffix resulted in the form önnön, which is first documented in late fourteenth century texts. The emphatic suffix was originally applied to each member of the personal pronoun paradigm, and the resulting complex forms functioned as emphatic possessors. However, this complex pronominal paradigm was gradually disappearing by the sixteenth century. Only önnön survived eventually, and it came to be used in all persons and numbers in possessive constructions as an emphatic marker.
3.2 Tulajdon ‘own’
As noted with respect to (19) above, tulajdon ‘own’ can replace saját ‘own’ in some possessive constructions salva veritate. Consider the following two examples from the Hungarian National Corpus for illustration:
(20) a. … de csak a tulajdon öreg arc-a nézett vissza rá!
but only the own old face-POSS.3SG looked back onto.him
‘… but it was his own old face that looked back at him.’
b. … hevesen udvarol-ni kezdett tulajdon feleség-é-nek.
passionately court-INF started own wife-POSS.3SG-DAT
‘…he started to court his own wife passionately.’
Tulajdon ‘own’ is nevertheless more constrained in its grammar than saját ‘own’. First, it only occurs in possessive constructions, functioning as an intensifier adjective, but it is ungrammatical in non-possessive constructions in the low position reserved for classificatory adjectives. I repeat (7a) as (21) to illustrate:
(21) Szó van egy új saját/*tulajdon lemez-ről …
word is a new own album-about
‘They are talking about a new album of their own …’
Second, tulajdon ‘own’ typically occurs in possessive constructions where the head noun is an inalienable possession of the possessor. So it is most natural if the possessum is a body part
noun (as in 20a), a kinship term (as in 20b), or it refers to a psychological property of the referent of the possessor. Saját ‘own’ is not constrained by such collocational restrictions, and it is free to co-occur with any noun that can be conceptualised as a possessum. The inalienable possession restriction that characterizes the grammar of tulajdon ‘own’ has been reported to exist for certain possessive adjectives in other languages, too (see, for example, iðios ‘own’ in Modern Greek, discussed in Alexiadou 2005).
In short, tulajdon ‘own’ is an alternative to the possessive saját ‘own’ in possessive structures where the head noun is an inalienable possession. Just like saját, tulajdon also precedes any other adjectives that modify the head (see 20a). It can also co-occur with a pronominal possessor, once again in parallel with the behaviour of saját (cf. 10b):
(22) Maga szegény - feleli nekem az én tulajdon feleség-em.
you poor tells DAT-1SG the I.NOM own wife-POSS.1SG
‘You poor thing - tells me my own wife.’
(22) is from the Hungarian National Corpus. It has a slight archaic or Biblical touch, like most other occurrences of the word tulajdon ‘own’. Tulajdon ‘own’ is a possessive intensifier adjective that acts as an alternative to saját ‘own’ in certain registers.
3.3 Önnön ‘own’
As we saw in 3.1, önnön used to be one of the emphatic pronominal possessors, but it gradually developed into a non-agreeing possessive intensifier with the loss of this paradigm.
While tulajdon ‘own’ only shows an observable restriction to certain registers, önnön ‘own’ is definitely archaic for most native speakers. It is used nevertheless in contemporary texts still.
The following three sentences are from the Hungarian National Corpus:
(23) a. … alá is írta önnön politikai halálos ítélet-é-t.
PRT too wrote own political death warrant-POSS.3SG-ACC
‘…he did sign his own political death warrant.’
b. … felméri önnön belső világ-á-t …
surveys own internal world-POSS.3SG-ACC
‘…he surveys his own internal world …’
c. … sikerült legyőzni önnön gyávaság-om-at … managed defeat-INF own cowardice-1SG-ACC
‘… I managed to defeat my own cowardice.’
(23a) and (23b) show third person agreement on the possessum, while the noun head is marked for a 1SG possessor in (23c). The agreement facts do not support a pronominal analysis of önnön ‘own’ in contemporary Hungarian. In accordance with the earlier discussion in 2.3, I assume that the possessive noun phrases in (23) each contain a pro-dropped pronominal argument, which is modified by önnön ‘own’.
I conclude that önnön ‘own’ is also a possessive intensifier adjective in contemporary Hungarian, that is, it has developed into one from a diachronic pronominal source. It only occurs in possessive noun phrases, just like tulajdon ‘own’. The possessive noun phrase that contains önnön ‘own’ typically has no definite article on the left edge (see 23). True pronominal possessors require the presence of a definite article (see 11a). And, finally, önnön
does occasionally co-occur with overt pronominal possessors, as the following example (source: http://zene.hu/cikkek/cikk.php?id=3178¤tPage=3) shows:
(24) …, de az én önnön másik fel-em a hit-em.
but the I.NOM own other half-POSS.1SG the faith-POSS.1SG
‘…, but my own other half is my faith.’
So to the extent that native speakers still use this marker, they treat it consistently as an intensifier possessive adjective, not as a pronoun.
4 Summary and outlook
In this paper, I have overviewed the grammar of three possessive intensifiers in Hungarian:
saját, tulajdon and önnön ‘own’. I have argued that their attested morphosyntactic properties are only consistent with an analysis in which they are treated as adjectives. They are only special in having been specialised for a functional role that König & Vezzosi (2008) refer to as intensifier. They modify the overt or covert possessor in this capacity, and they act as two- place predicates identifying a pronounced possessive relation between the possessor and the possessum. In addition to this use, saját ‘own’ can also occur in non-possessive constructions, where it is inserted as a low-level classificatory adjective.
My fundamental aim in this article has been to lay down the basic facts that are needed to develop a descriptively adequate account of the grammar of possessive adjectives in Hungarian. These elements have not received much attention in the generative literature on Hungarian, and traditional descriptive grammars only discuss them in passim. I have shown that these three elements are adjectives, and they are not pronominal in nature. This is important since qua adjectives, they do not directly participate in the establishment of referential dependencies, a claim that I have tried to substantiate by showing that they are not subject to a strict grammatical constraint governing the choice of an antecedent. Nevertheless, the presence or absence of these intensifier adjectives does have an influence on how successfully the pronominal possessor can be bound. This influence, however, is not strictly grammatical in nature. My aim has been to establish this fact in this article, and I leave the study of how the presence of possessive adjectives influences the acceptability of specific binding configurations to another occasion.
References
Alexiadou, A. (2005): A possessive adjective in the Greek DP. In: Stavrou, M. & Terzi, A.
(eds.): Advances in Greek generative syntax. In honor of Dimitra Theophanopoulou- Kontou. Amsterdam & Philadelphia: John Benjamins, 127-152.
Baker, C.L. (1995): Contrast, discourse prominence, and intensification, with special reference to locally free reflexives in British English. Language 71, 63-101.
Bárczi, G. (1941): Magyar szófejtő szótár Hungarian Etymological Dictionary. Budapest:
Királyi Magyar Egyetemi Nyomda.
Chomsky, N. (1986): Knowledge of Language. New York: Praeger.
Davies, M. (2004-): BYU-BNC. (Based on the British National Corpus from Oxford University Press). Available online at http://corpus.byu.edu/bnc/.
Fiengo, R. & Higginbotham, J. (1981): Opacity in NP. Linguistic Analysis 17, 395-421.
Hadrovics, L. (1998): Hátrahagyott szófejtések Etymological inquiries left behind for the posterity. Magyar Nyelv XCIV (4), 398-406.
Keszler, B. (ed.) (2000): Magyar grammatika Hungarian Grammar. Budapest: Nemzeti Tan- könyvkiadó.
É. Kiss, K. (2002): The syntax of Hungarian. Cambridge: Cambridge University Press.
König, E. & Vezzosi, L. (2008): Possessive adjectives as a source of intensifiers. In: Seoane, E. & López-Couso, M.J. (eds.): Theoretical and empirical issues in grammaticalization.
Tpyological studies in language, Volume 77. Amsterdam & Philadelphia: John Benjamins, 183-206.
Laczkó, T. (1995): The Syntax of Hungarian Noun Phrases: A Lexical-Functional Approach.
Frankfurt am Main: Peter Lang.
D. Mátai M. (1999): A névmások történeti a középmagyar kor végéig The history of pronominals till the end of the Middle Hungarian Period. Magyar Nyelvőr 13(4), 438-464.
Nishiguchi, S. (2009): One's own. Talk delivered at ESJ 2nd Internatioanl Spring Forum on 26 April, 2009. Nara Women's University. Accessed at http://www.rs.tus.ac.jp/ nishiguchi/
ELSJslides.pdf
Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J. (1985): A comprehensive grammar of the English language. London: Longman.
Rákosi, Gy. (2011): Összetett visszaható névmások a magyarban Complex reflexive pronominals in Hungarian. In: Bartos, H. (szerk.): Általános Nyelvészeti Tanulmányok XXIII. Budapest: Akadémiai Kiadó, 351-376.
Rákosi, Gy. (2014): Possessed by something out there: On anaphoric possessors in Hungarian. Argumentum 10. Debrecen: Debreceni Egyetemi Kiadó, 533-547.
Reinhart, T. (1983): Coreference and bound anaphora: A restatement of the anaphora questions. Linguistics and Philosophy 6, 47-88.
Safir, K. (1996): Semantic atoms of anaphora. Natural Language and Linguistic Theory 14, 545-589.
Szabolcsi, A. (1994): The Noun Phrase. In: Kiefer, F. & É. Kiss, K. (eds.): The syntactic structure of Hungarian. Syntax and Semantics 27. New York: Academic Press, 179-275.
Váradi, T. (2002): The Hungarian National Corpus. In: Proceedings of the 3rd LREC Conference. Las Palmas, 385-389. http://corpus.nytud.hu/mnsz.
Zaicz, G. (szerk.) (2006): Etimológiai Szótár Etymological dictionary. Budapest: Tinta Könyvkiadó.
Zribi-Hertz, A. (1995): Emphatic or reflexive. On the endophoric character of French lui- même and similar complex pronouns. Journal of Linguistics 31, 33-374.
György Rákosi University of Debrecen
Institute of English and American Studies H-4010 Debrecen
Pf. 73
rakosigy@hotmail.com rakosi.gyorgy@arts.unideb.hu
